realists, constructionists, and lemmings oh my! (part II)
In my last post, I argued that in order to understand systems of valuation, we need to develop an intermediate position between realism and constructionism. In particular, I said “we need an intermediate position… that reconciles:
(a) The idea that prices (or more generally, the prevailing valuation of anything including, e.g., one scientific theory vs. another) reflect subjective valuations and these valuations are governed, in the first instance, by (the anticipation of) others’ subjective valuations; with
(b) The idea that such valuations are constrained by fundamental and objective values.
Moreover, such a position should hopefully be able to explain why those constraints operate so loosely for periods and then tighten up. And this question is particularly vexing since (in the case of the housing/subprime bubble), we clearly have a case where there was a massive bubble but there was not, contrary to the talking heads quoted [in my previous post], a mass delusion. What did people believe then, and what changed?
Before outlining such an approach in the case of financial markets (and the stock market, in particular), let me underline the fact that the imperative of reconciling (inter)subjective, constructionist processes with real, objective constraints is a general one that pertains to any system of valuation. Abbott is very articulate about this in the System of Professions (1988: 35-40), where the question at hand is to reconcile two views concerning the fact that there are often many and inconsistent ways of defining a professional task or problem. For example, the “problem” of alcohol has variously been regarded (by professions intent on enlarging their domains) as a medical problem, a criminal problem, a moral/spiritual problem, etc. In short, “Betwewn the objective facts of alcohol consumption… and the subjective definition of alcoholism qua problem there is thus a long and tenuous relation (p.37).” What to do?
There are two easy, and erroneous, views of this relation. The first holds that problems are given facts and that different perspectives on them are errors or ideologies. The second argues that problems have no objective reality whatsoever and are the simple creations of self-seeking groups. Neither view works. To deny the objective features of the alcohol problem is silly. The objective features have changed little, although increased body size and health may have increased tolerance to the drug. But it is just as silly to claim that these objective qualities determine (or should determine) the subjective perception of the problem. Alcoholism has been a biological disease, a personal problem, a moral delict, a sin. It has even been a “scourge of society”– a problem caused by poverty and injustice and hence not really an individual’s problem at all. Yet all these reinterpretations and reconstructions were bounded by the objective properties of alcohol consumption itself (pp. 37-38; emphasis added).
In short, the challenge is to how objective/real features of the target of (classification and) valuation give shape to how those targets are (classified and) valued. This applies in any domain of valuation. For instance, while it is not fashionable to say so, it is obvious that objective features of the physical world constrain the theories we see in the natural sciences, and the same is true for how objective aspects of human beings and how they interact constrain social scientific theories.
Of course, such constraints will operate only insofar as the incentives are in place for such scientists to care about having valid accounts. Unfortunately (for people like me, who believe in scientific progress), that is not always the case. Rather, a premium is placed aesthetic desiderata (e.g., the “sexiness” of a topic; the felicitous coining of a phrase; its political associations). This is clearly the case in academic communities where the terms “reality” and “objective” are dirty words. Though it is important to note that even aesthetic efforts (in such fields) face objective constraints in the form of legitimized modes of discourse within which they must fit.
This may seem like a strange usage of the term “objective,” so let me transition to the case of financial markets by drawing on Abbott for how to distinguish subjective and objective and how the latter constrains the former:
First, the difference between objective and subjective is quite clear in the case of alcoholism:
Through all this (reconstruction) alcoholism always retained certain objective qualities. It always involved individual consumption of alcohol. Consumption always produced central nervous system depression with loss of fine, and ultimately, coarse motor skills (p.37).”
A key implication of this is that while there has been a lot of variation in how alcoholism has been defined, there has been relatively little overlap (with some exceptions; e.g., Mormon restrictions) in how caffeine has been regarded. Different drugs–>different physiological effects–> different spaces available for social construction.
Second, this approach does not apply only to cases where the target of valuation is something physiological. It applies even when the target is cultural/constructed. The key is the relative mutability/inertia of (particular features of) the target:
The real difference between the objective and subjective qualities of problems is a difference in amenability to cultural work. In the preceding chapter, I argued that professions sometimes use their abstract knowledge to reduce the work of competitors to a version of their own. This is a basic mechanism of interprofessional competition. The objective qualities of a task are those that resist such reconstruction. Despite the normal connotations of the word, objective qualities are not a reality that awaits discovery beneath the cultural images; they are an inertia that reconstruction must overcome (pp. 36-37; emphasis added).
[For instance,] the idea that alcohol consumption is an individual rather than a group activity is a group activity is a cultural belief, not a natural fact. But for the purpose of most jurisdictional claims, such deep-rooted beliefs have an objective quality. They may change as the culture changes, but jurisdictional claims are made over a considerably shorter time-scale, within which these cultural facts appear like objective facts. The opposition between the objective and subjective in human problems is not between the natural and the mental, but between the movable and the fixed (pp. 38-39; emphasis added).
Swidler, in Talk of Love, fleshes this distinction out very nicely. In particular, while the institution of monogamous marriage may be a construction, it is so institutionalized that it confronts each of us as a relatively objective feature of our reality (though, its gender component is changing in our generation!), presenting us with challenges that we address via cultural work.
Finally, Abbott suggests how it is that objective features of a task shape construction:
(While) there is no easy formulation for this bounding of subjective by objective dimensions, (t)he further a reinterpretation takes a phenomenon from its objective formulation (e.g., it would be a real stretch to claim that the challenge in having a good marriage is to cope with the fact that daughters-in-law and mothers-in-law get along too well!), the more energy that reinterpretation requires. While this energy comes originally from the redefining profession, it comes later from public acceptance. Reinterpretations are normally part of larger jurisdictional claims, claims not only to classify and reason about a problem, but also to take effective action towards it. The final tests of such claims are their practical results, whatever the criteria for efficacy may be in a given culture. Since public acceptance ultimately requires results, it follows that the greater the deviation from objective qualities, the more necessary are measurable results (p.38; emphasis and in-laws example added).
So, constructionist projects (i.e., any attempt at valuation) are limited by the practical results by those who pursue them. They need to be efficacious in navigating the objective features of the target of valuation, where objectivity is defined as above.
Let’s port that into the case of stock market valuation by asking two sets of questions:
1. Did you think internet stocks were too high in 1999? And if you did, what did you do about it? And the same goes for, say, California housing prices in 2005: did you think they were too high? And if so, what did you do about it? And if you said they were too high but you did nothing about it, what would have to change about the world so that you might actually do something about it?
2. What would you do if you could buy General Electric for $100 but you were told you could never sell it (or any of its divisions, etc.)? Alternatively, what if you were told you could buy an apartment building on the East Side of Manhattan for $100 but told you could never sell it (or go condo, co-op, etc.)? Finally, if you said that you would buy these at $100, how high a price would you be willing to pay?
I’m going to give my answers to these questions now, but I would really encourage the interested reader to try to answer them on your own before hearing my answers.
Ok, here they are. Let’s start with the easiest answer– no one would turn down the opportunity to buy GE or a Manhattan apartment building for $100. The important thing is to be clear about why that is. One possible reason is that you could turn around and resell them for much much more money. But pay attention: I have taken away that possibility by saying you must own them forever. But the answer remains the same. Why is that? The answer, quite obviously, is that these assets generate an income stream that is worth much more than $100. Even in the first month, you would receive income far more than $100. (In the case of GE, you wouldn’t even need to generate additional income beyond the cash GE has in the bank). And you would receive such >$100/mo. far into the future. Note that implicit in all this is that ownership has objective features (where objectivity is defined as above)– it conveys the right to the income from the asset owned. Again, obvious but important. This leads us to the key lesson: The lower bound on the price of ownership is the value of the expected income stream from the asset. Let’s call this, its fundamental value. In the case of GE, these are the profits from its divisions; in the case of an apartment building, it is the rental income [which is true for all residential real estate– if for whatever reason, you could rent your house for $1m a month, you would move out and do that].
Now, it is reasonable at this point to become wary– How can an “expectation” be objective?? Fair enough. But pay attention to two points:
(a) I have set the price at $100. All I’m saying is that the price will not get that low because the value of the expected income from these assets is much greater than $100. True, we live in a constructed world. It is possible that, in between the time that you take ownership of GE and the time you go to the bank and withdraw its funds, Obama might really turn out to be a Commie and would take away your right to these funds; we can think of a zillion possible but highly-unlikely scenarios of this type. But in practice we don’t. We take the reality of our institutions for granted because they just don’t change that fast, in part because there are powerful interests that are invested in the status quo. So in practice, there is in fact an objective lower-bound or common denominator that we all can agree is true for everyone. Like gravity.
(b) In the case of financial markets at least, we don’t even have to agree about the lower-bound for it to function as a floor. All that is necessary is that one person, who either has $100 or can borrow it from his mother-in-law, believes that these assets are worth more than a particular price for this to be a lower bound. Once the price hits this level, we can expect the person to immediately buy the asset. After all, she incurs no risk in doing so [other than that Obama is a Commie]. She does not have to worry about what other people believe and whether she can resell it to them. She just needs to believe that if she were to own the asset, she would gain ownership of an income stream than is worth more than the current price. And note crucially that this person may in fact be unreasonably optimistic. Perhaps she is convinced that GE is worth $300B (its current market value is $191b). If she can get access to such funds (I will return to this!), she will buy it.
These considerations are the key ingredients in understanding where asset prices come from. Fundamental value, as defined by the most optimistic investor (who has access to enough capital to buy controlling interest), establishes a floor on such prices. Where exactly is that floor for each asset? It is impossible to say. But we can say for sure that say, the price of GE stock will not go under an implied market value of $100. All that is necessary to believe this is is to believe that at least one person out there believes that this price is below its fundamental value. It’s kind of like saying that we don’t believe that $100 would last long lying in Times Square. We might not pick it up. But we certainly expect someone to do so. It is as predictable as gravity.
Now in practice, most stock-market investors neither seek, nor have the wherewithal, to seek controlling interest in a company. But again, all that is necessary to set this lower bound is that someone does (and note that, as the price declines, the capital required for taking it private declines too– thus hardening that floor on prices). For instance, Warren Buffett. This is his main mode of operation. And I can tell you right now that every stock that goes lower than its expected income stream, as he sees it, is at risk of being taken private by him. And of course, he is not alone.
Ok, now let’s focus on the first question– if you believed, as I, and (much more importantly) Warren Buffett did, that prices (for internet stocks in the late 1990s; for residential real estate in the 2000s) were way above fundamental values, why didn’t we do anything about it? One interpretation, as given by the talking heads discussed in the last post, is that in fact we had all deluded ourselves into believing that prices were right. But that is wrong since, as I demonstrated, there was much public discussion of the possibility that we were in a bubble. A lot of people doubted such prices but apparently did nothing about it. Why?
The reason is that there was basically nothing they could do. As discussed above, prices will not go below fundamental value, as defined by the most optimistic, well-capitalized investor. But what can the well-capitalized, pessimist do if the optimists drive up the price far in excess of her estimate of fundamental value? In the case of stocks, the only option is to short the stock– i.e., borrow shares and sell them, and then buy them back later if the price goes down. That is, you take profit from selling low and buying high. This can work, and short-sellers can make a lot of money from doing it. And this activity does in fact help to prevent prices from getting too high. But it is crucial to recognize that this pessimistic ceiling on prices relies on actors and institutions, with accompanying risks, has no counterpart when it comes to the optimistic floor on prices discussed above. To repeat, the floor on prices comes from the most optimistic, well-capitalized investor’s belief that the current price is below her estimate of its fundamental value. But the ceiling on prices depends on much more than the well-capitalized investor’s belief that the price is above the asset’s fundamental value. In particular, it must also be true that; (a) the rate of interest on the loan of shares is low enough to justify the loan; (b) she will be able to buy back shares when the prices declines; and most crucially, (c) the price really will decline. The last point is the most crucial because it tells you that the ceiling on stock prices depends on what other investors will come to beleive– it is a speculative maneuver, as defined by Keynes. And speculation is always risky– what if the optimistic herd keeps on growing? Oy veh. You will have to buy back the stock at a higher price– i.e., cover your short position, and thereby lose your shirt [e.g., you will have borrowed [and then resold] at $100, but then paid someone else $200 to repay that loan]. Moreover, what if there are only a few shares available when you want to? Oy veh again. Since supply is limited, you and your fellow shorts will be fiercely competing to get those shares, thereby driving up the price you will have to pay. This is exactly what happened to a whole bunch of woe-begotten hedge funds (boo hoo) who had shorted Volkswagen but got caught in a short squeeze [where the supply of shares was limited, and perhaps even monopolized or cornered], thereby causing them to lose a gazillion dollars. In sum, there are two key lessons here:
(a) the floor on prices is much harder than the ceiling; and
(b) insofar as this ceiling is made of straw rather than brick, price increases are driven do not reflect a consensus, but a minority view (see esp. Miller 1977).
Ok, now let’s apply this in the case of internet stocks and recent residential real estate prices. For the former, the key point is again the supply of shares or “float” of such stocks. As discussed above, even a pessimist who is absolutely convinced that prices are about to drop– perhaps especially because he believes that there are a lot of other people who agree with him– will worry if the supply of shares is limited. If prices go up in the short-term, he will lose his shirt. And in fact, there is very good evidence that short-selling of interet stocks was very difficult in the late 1990s because of their very limited float, thereby limiting short-selling, and letting the minority of optimists drive prices. For more on this, see e.g., Ofek & Richardson 2003 and Hong, Scheinkman, and Xiong 2006, both of which are based on the ideas set forth by the Miller article above and are specific versions of the general point that there are limits to arbitrage that prevent prices from equaling their fundamental values. This idea, which as I have said, hearkens back to Keynes— has gotten an incredible amount of traction in recent years in the finance community [cites to Miller 1977 through 1986: 23; cites from 1987-1996: 36; cites from 1997-2006: 92; cites from 2007-2008: 41; that’s the pattern I want my articles, though I don’t particularly want Miller’s renown in other respects]. What seems less commonly recognized in the finance literature (and is not clarified by Keynes) is that these limits do not affect the floor on prices, just the ceiling.
Finally, let’s turn to real estate prices. Here the issue of the ceiling-of-straw is even clearer. Until very recently, it simply was not possible to short residential real estate. The institution did not exist. This meant that while prices would not go lower than the most optimistic investor’s estimate of rental income, prices were essentially unlimited on the upside because the ceiling was made of straw. And again, under such conditions, it is a minority view that drives prices. Note that one might argue that one does not need short-selling to be the ceiling in such markets because real estate investors borrow their money and so prices are limited by they were limited by banks’ beliefs about the value of the real estate, which they take as collateral. But as discussed in an earlier post, various interrelated trends combined to lead lenders to effectively abdicate what was in their self-interest. (Note that the other fundamental constraint on real estate prices besides rents is household income because lenders will not lend to borrowers who cannot pay back… but they decided to forget about that too). And just as the increased float of internet stocks, due to the expiration of lock-up agreements on insiders’ shares, allowed (selling) and short-selling to occur and thereby pessimistic views to deflate internet stocks, the U.S. residential real estate price bubble seems to have been pricked by the emergence of a vehicle by which investors could effectively short subprime, via the ABX indexes. As Gorton [must-reading, for anyone who wants to understand this crisis] argues:
… the ABX indices as reveal[ed] hitherto unknown information, namely, the aggregated view that subprime was worth significantly less. In fact, some of the dealer banks themselves, we now know, were shorting the index to hedge their long positions—of course so was everyone else. The ABX indices also allow all parties, e.g., hedge funds, to express their views on the value of subprime RMBS bonds. It is not clear whether the housing price bubble was burst by the ability to short the subprime housing market or whether house prices were going down and the implications of this were aggregated and revealed by the ABX indices. It seems that the indices played a central informational role (2008: p.23).
Regardless of how important a causal role one gives to these indexes (see p.23 of his paper for a graph of the sharp decline in these indexes and when that decline occurred), Gorton’s key point is that prior to their introduction, any pessimism on subprime simply could not be expressed. There was no vehicle for such expression. And the related point is that again, the fact that prices were high did not mean that everyone had drunk the Kool Aid. In fact, as Gorton emphasizes, the very people who were selling the Kool Aid were worried it was toxic, and so when they had an opportunity to express their view [thereby hedging their bets], they did so. But since it turned out that many people had been harboring such pessimistic views, they drove that index down; and since this told everyone that such pessimism was widespread, all of a sudden everyone realized that they needed to head for the exits. Oops. Turns out there were a lot of people and the door was closing. And that ends up sucking for all of us.
Note crucially that for the most part, this was not a story where the actors involved really were sharply distinguished between optimists (100% belief that prices were always going up) and pessimists (100% sure that prices were insane). Rather, optimism and pessimism typically compete within the mind of each person. Let’s say that the various miscreants participants in the housing or internet bubbles were actually on the fence– 50% that it was a bubble, 50% that prices were reasonable. Many such people might still have acted as if they were optimists, especially if the two mechanisms discussed in my earlier post were operative– i.e., they believed that the prevailing view was optimistic (which was reasonable because pessimistic voices could not be expressed) and especially if they were agents rather than principals (thereby creating a condition where they were worried about going out on a limb as the lone pessimist). But the fact that people’s beliefs were more pessimistic than their action is a big part of the reason why the rush to the exit was so rapid– The interpretation was not “Oh shit. I guess I was wrong all along!” Rather, it was: “Oh shit. It was exactly as I feared all along!”)
Phew. This is an insanely long post. I will actually pick up a theme from this last bit on subprime in my final post for orgtheory, which will be on why social networks are overrated (heresy for me, given my academic upbringing). For now, let me conclude by summarizing where we are with respect to the general agenda outlined at the outset of this post:
1. The target of valuation (e.g., the physical world for science; a personal or social problem for professions; the cash generated by a financial asset) constrains valuation because there are objective limits to human experience with the target. This is Abbott’s “practical test” as above. And to repeat (based on Abbott), objectivity here means simply that there are features of the target that are slow to change relative to the time-scale that pertains to its valuation.
2. Such constraints on valuation will be looser or tighter depending on whether those who are more creative in their constructions–“optimists”– must contend with those who hew closely to the lowest common denominator on what is objective– “pessimists.” If pessimistic views cannot be expressed, either because the institutional vehicle for such expression simply does not exist or because use of that vehicle is risky, the optimists are constrained only by what they can convince each other. Castles can thereby be built in the air. And the constraints on pessimism/realism can cause the optimists to fool themselves into thinking that the castle is built on solid ground, at least for a time. But when the pessimism is able to be expressed, all of a sudden everyone realizes that they were in the air after all.
3. The shadow of realist constraints can be traced in any constructionist project. What do I mean by this? Well, note that the last line in the last paragraph may seem glib. All because the agents of realism– pessimists– can express their views, does not mean that the agents of construction– optimists– have to give up their hopes and dreams. Why can’t they agree to disagree? They can, but not for long. Why? Because the constructionist projects are built on successful coordination. They work to create a reality only insofar as actors can be confident in that reality and its staying-power. This is not a problem if the construction does not stray far from an interpretation that any member of a system can experience as objective reality. But insofar as it involves interpretations of that reality that are more contestable/debatable (i.e., that involve abstractions from the common-denominator reality), it will become difficult to sustain coordination as each actor is less confident that other actors will do likewise. This is what Abbott meant in the quote above about the “energy that reinterpetation requires”). So the silence of pessimism is crucial here. As long as it cannot be expressed, optimists can think that optimism is the consensual view. (A key Talmudical principle is “Shtika ke’hoda’a”– Silence implies acceptance. But this doesn’t apply if there is tape over the would-be objector’s mouth; the problem is that tape is often hard to see). But once pessimism is expressed, coordination is undermined. And insofar as there is a threat that the noise level will rise as the construction moves away from objective reality, this tells us a lot about the nature of constructionist projects. In particular, they are couched in realist/objective terms. Optimists do not simply think that prices should be higher than pessimists do. They articulate and invoke a theory of value that justifies their price estimate via an interpretation of fundamentals. Thus, constructionist projects are bounded by their reasonableness as interpretations of the objective world. This does not mean that there are not constructions that stray far from objective values. But in general and over time, the more such theories abstract away the objective features that everyone can agree with, the less reasonable they will seem, and those that at first seem reasonable are vulnerable to being shown to be wrong. To paraphrase Abbott again, any constructionist projects must be “efficac(ious).”
In sum, it’s not that social constructions can stray far from reality, it’s just that this straying is much harder when more realist voices can express themselves, and that while constructionist projects have an easier time getting off the ground if they seem like reasonable interpretations of fundamentals, they are then subject to the “practical test” (Abbott, above) via “inspection and evaluation.”
A final note: I have enjoyed orgtheory and will miss doing this after my time is up next week. But as I have publicly fretted, I just don’t have the time for this. Also, I think some of the back-and-forth in the comments has not been particularly productive. So… if you choose to comment on this or any of my final posts, I will only respond if I think I have something to say in response that can lead to a productive exchange. So please do not infer anything from my silence (generally a good rule to follow!)– and I will certainly not infer anything from anyone else’s silence. But if anyone wants to engage with me further on anything and I don’t respond, please just email me.